Summary of a Health Effects Institute (HEI) Special Report
HEI Diesel Working Group
Diesel engine emissions are highly complex mixtures. They consist of a wide range of organic and inorganic compounds distributed among the gaseous and particulate phases. Public health concern has arisen about these emissions for these reasons:
The particles in diesel emissions are very small (90% are less than 1µm by mass), making them readily respirable.
These particles have hundreds of chemicals adsorbed onto their surfaces, including many known or suspected mutagens and carcinogens.
The gaseous phase contains many irritants and toxic chemicals.
Oxides of nitrogen, which are ozone precursors, are among the combustion products in the gaseous phase.
There is a likelihood for humans to be exposed to diesel emissions or their atmospheric transformation products in both ambient and occupational settings.
Diesel emissions have the potential to cause adverse health effects. These effects include cancer and other pulmonary and cardiovascular diseases. However, diesel engines are only one of many sources of ambient particulate matter and gaseous air pollutants. Therefore, it is difficult to measure the exposures from various sources, and to distinguish the potential health risks attributable to exposure to diesel exhaust from those attributable to other air pollutants.
For over a decade, the HEI has supported a broad-based research program to evaluate the health risks of diesel emissions, including investigations of carcinogenesis, modeling studies, and emissions characterization. It also organized a Diesel Working Group consisting of scientists with expertise in automotive engineering, atmospheric chemistry, toxicology, pathology, molecular biology, epidemiology, and environmental sciences to examine what is known, not known, and still uncertain about the health risks of exposure to diesel emissions.
The HEI Diesel Working Group focused its evaluation on a set of issues that it thought were critical to assessing the carcinogenic risks of exposure to diesel exhaust. Its Special Report1 includes background papers that contain in-depth discussions of emissions, exposure, toxicity, carcinogenicity, and dose-response relations. The Report also contains a paper that synthesizes the Working Group's determination of what conclusions about the carcinogenicity of diesel exhaust could be drawn from the available scientific data and identifies important information gaps.2 The major findings of the HEI Diesel Working Group are discussed in this summary.
The composition of diesel exhaust varies considerably depending on engine type and operating conditions, fuel, lubricating oil, and whether an emissions control system is present. Diesel engine emissions have changed dramatically over the last 30 years because of improvements in engine technology, emissions controls, and fuel formulation. Emissions of oxides of nitrogen and particulate matter from the diesel engines introduced in the late 1980s and early 1990s are significantly lower than those from older engines. As a result, characterizations of modern-day diesel exhaust cannot be used to estimate past exposures, nor can they be used reliably to project future emission profiles.
It is very difficult to assess exposure to diesel emissions because they are highly complex mixtures and constitute only a small portion of a broader mix of air pollutants. For example, combustion of other materials, such as fossil fuels and tobacco, produce many of the same chemical components that are present in diesel emissions; furthermore, both natural and man-made sources of respirable particles are common. No single constituent of diesel exhaust serves as a unique marker of exposure; however, scientists can use the levels of fine particles or elemental carbon (both of which are much higher in diesel emissions than in other combustion products) as surrogate indices of diesel exhaust particulate matter. When estimating exposure to diesel emissions, the following factors need to be considered:
Because of improvements in engine design and emissions control technology, and the use of reformulated fuels, future human exposures to diesel engine emissions will differ from past and current exposures. However, reductions in exposure to diesel emissions will be gradual because of the long life of heavy-duty diesel engines, and will be offset as the use of diesel engines grows.
The fact that the chemical and physical characteristics of diesel emissions will change as new technology and fuels are implemented cautions against automatically assuming that a decrease in the amount of emissions will result in a decrease in risk.
Once emitted, diesel emissions undergo atmospheric transport and transformation processes that may alter the toxic, mutagenic, or carcinogenic properties of the original constituents, creating new products that may be either more or less hazardous than the original emissions.
Exposure to diesel exhaust particulate matter has been assessed in occupational settings and some ambient environments. Although the existing data are limited, some estimates of the range of human exposure to diesel emissions can be made:
In some occupations, diesel emissions contribute a high proportion of the particulate and gaseous air pollutants. The estimates for workplace exposures to diesel exhaust particulate matter range widely, from approximately 1 to 100µg/m3 (eight-hour averages) in some occupations such as trucking or transportation, to 100 to 1,700 µg/m3 for occupations such as underground mining where equipment powered by diesel engines is often used in enclosed spaces.
The information on ambient exposures is sparse. In an analysis conducted in the Los Angeles Basin in the early 1980s, diesel emissions accounted for approximately 3% of the mass of total particulate matter and 7% of the mass of fine particles emitted into the atmosphere. Average monthly values for ambient levels of diesel exhaust particulate matter ranged from 1 to 3 µg/m3 in areas with low levels of air pollution. These values are in general agreement with the range of nationwide annual average values derived by the U.S. Environmental Protection Agency using vehicle emissions factors, sales information, and pollutant exposure models. In the Los Angeles study, the highest monthly average levels of diesel particulate matter were approximately 10 µg/m3 at the most polluted locations during winter months, the period of highest exposures. Short-term or peak exposures to diesel particulate matter, especially in urban settings such as street canyons, are usually higher than monthly or annual average concentrations.
Given the limited exposure information, it is a challenge to determine the contribution of diesel exhaust to human cancer. The Diesel Working Group developed the following conclusions after reviewing over 30 epidemiologic studies of workers exposed to diesel emissions in occupational settings for the period 1950 through the early 1980s.
The epidemiologic data are consistent in showing weak associations between exposure to diesel exhaust and lung cancer. The available evidence suggests that long-term exposure to diesel exhaust in a variety of occupational circumstances is associated with a 1.2- to 1.5-fold increase in the relative risk of lung cancer compared with workers classified as unexposed.
Despite the concern that confounding by cigarette smoke might explain the observed risk elevations, most studies that controlled for smoking found that the association between increased risk of lung cancer and exposure to diesel emissions persisted after such controls were applied, although in some cases, the excess risk was lower. Only a few epidemiologic studies considered other potential confounders such as nondiesel particles, environmental tobacco smoke, asbestos exposure, diet, and socioeconomic factors. At present, there is insufficient evidence to conclude whether confounding by these factors influenced the results.
As is frequently the case in epidemiologic studies of air pollutants, none of the studies measured exposure to diesel emissions or characterized the actual emissions from the source of exposure for the period of time most relevant to the development of lung cancer. Most investigators classified exposure on the basis of work histories reported by the subjects or their next of kin, or by retirement records. Although these data provide relative rankings of exposure, the absence of concurrent exposure information is the key factor that limits interpreting the epidemiologic findings and using them to make quantitative estimates of cancer risks.
The carcinogenic activity of diesel emissions has been convincingly demonstrated in rats. Nearly lifetime exposure for 35 hours or more per week to high concentrations of diesel exhaust particulate matter (2,000 to 10,000 µg/m3) causes an exposure-dependent increase in the incidence of benign and malignant lung tumors in rats. No consistent evidence suggests that diesel emissions induce cancer in rats at sites other than the lung. Prolonged exposure to diesel emissions does not produce lung tumors in hamsters, and the results in mice are equivocal, which suggests that species-specific factors play a critical role in the induction of lung tumors by diesel emissions.
Recent reports from two independent laboratories support the idea that the particle-associated organic chemicals play little or no role in the development of lung tumors in rats exposed to high concentrations of diesel emissions. No significant differences were noted in tumor incidence or histopathologic characteristics between rats exposed to diesel exhaust and those exposed to carbon black (a surrogate for the diesel particles minus the adsorbed organic compounds). These results do not completely eliminate a possible role for the adsorbed chemicals, some of which are potent mutagens and carcinogens. If bioavailable, they could play a role in carcinogenesis that might not be detectable in the rat bioassay because their effect is either too subtle or is masked by the overwhelming response of the rat's lungs to high concentrations of inhaled particles.
Even though the evidence strongly suggests that prolonged exposure to high concentrations of diesel exhaust particulate matter induces lung tumors in rats, the Diesel Working Group recommends caution in extrapolating these results to humans for the following reasons:
The lung tumors observed in rats exposed to high concentrations of diesel emissions may be due to a species-specific response to inhaled particulate matter rather than to a carcinogenic mechanism that also occurs in humans. When rats and other laboratory animals are exposed to high concentrations of diesel exhaust particulate matter or other poorly soluble nonfibrous particles for long time periods, lung clearance mechanisms are impaired and particles gradually accumulate in the lungs; this condition is referred to as lung overload. In the rat, lung overload has a characteristic threshold and initiates a progressive series of cellular responses, including inflammation, alveolar epithelial cell proliferation, and fibrosis. These responses are more severe in rats than in mice or hamsters, and appear to be associated with the subsequent development of lung tumors.
Although characteristic exposure thresholds for lung overload, as well as for the nonneoplastic and neoplastic responses, have been noted in the rat, extrapolation of no-effect levels for exposure to diesel exhaust from one species to another is problematic because of wide inter- and intraspecies variations in particle clearance rates and in susceptibility to cancer.
Our knowledge of the mechanisms by which prolonged exposure to high concentrations of diesel emissions produces lung tumors in rats is incomplete. At the high concentrations of diesel emissions used in the rat bioassay, the data imply that the diesel exhaust particulate matter triggers inflammation and cell proliferation. Such responses are thought by many scientists to cause cancer through indirect or "nongenotoxic" mechanisms rather than by direct interaction with DNA, as would be caused by the mutagenic chemicals adsorbed to the particles. At this time, however, only circumstantial evidence supports the hypothesis that diesel emissions induce rat lung tumors by nongenotoxic mechanisms.
The rat bioassay data do not exclude the possibility that diesel exhaust may induce lung cancer by different mechanisms in different species, or by different mechanisms in the same species at different exposure levels (e.g., predominantly nongenotoxic mechanisms under high-dose exposure conditions and genotoxic mechanisms under low-dose exposure conditions).
The Diesel Working Group cautioned that using the rat bioassay data (obtained at high-dose exposure levels) to make quantitative estimates of the carcinogenic risk of exposure to diesel emissions at environmentally relevant exposure concentrations may overestimate risk if the mathematical models used to extrapolate from high to low doses and from animals to humans do not (1) account for particle overload and associated inflammatory and proliferative processes, (2) recognize the apparent existence of a threshold for particle-induced biologic responses, such as impairment of lung clearance mechanisms, inflammation, cell proliferation, and tumor development, and (3) consider the mechanistic relation of the nongenotoxic injuries to the development of lung tumors in laboratory rats.
The Diesel Working Group found that it is not presently possible to base a risk characterization of diesel exhaust solely on either the human or the animal data. Instead, the Working Group evaluated and integrated the available information from diverse data sets to make the most informed judgments about the potential carcinogenicity of exposure to diesel exhaust2.
Key issues concerning the human health risk of diesel exhaust are: Does particle overloading occur in humans under environmental exposure conditions, and if so, does it trigger processes that lead to lung cancer. In the rat, the animal species most sensitive to diesel exhaust, lung tumors are produced after nearly lifetime exposures for 35 hours or more per week to high concentrations of diesel exhaust particulate matter (2,000 to 10,000 µg/m3). These concentrations are approximately three orders of magnitude higher than current estimates of average atmospheric concentrations of diesel exhaust particulate matter (1 to 10 µg/m3). One mathematical extrapolation model suggests that lung clearance mechanisms would not be impaired in humans even if they were exposed continuously (24 hours per day) to levels of particulate matter in this ambient range. According to this model, the levels of respirable particles that would be needed to depress lung clearance mechanisms in humans under continuous exposure conditions are greater than 100 to 200 µg/m3. This, however, is an unlikely exposure scenario, even for most workers. Under more realistic intermittent exposure conditions (eight hours per day, five days per week), the model predicts that the concentration of particulate matter needed to impair lung clearance would be 500 to 1,000 µg/m3. Only a limited number of workers, primarily miners, are exposed to concentrations of diesel exhaust particulate matter close to this range.
If we assume that particle-induced mechanisms of lung tumorigenesis operate similarly in rats and humans, the analysis above implies that there is some biological rationale for extrapolating the rat bioassay data to the small population of workers who are routinely exposed to high concentrations (greater than 1,000 µg/m3) of diesel exhaust particulate matter and who may have impaired lung clearance mechanisms. Because of the large interspecies differences in particle clearance, the rat bioassay data also may be relevant to those workers who are exposed to levels of diesel particulate matter one order of magnitude lower (100 to 1,000 µg/m3). However, the toxicity and modeling data do not support the assumption that exposure to diesel exhaust particulate matter alone at the levels found in most ambient settings (1 to 10 µg/m3) would be sufficiently high to overwhelm lung clearance processes and, thus, induce lung tumors by a mechanism driven by inflammation and cell proliferation.
A wealth of information is available about the potential for diesel emissions to cause cancer. Epidemiologic studies of different occupational cohorts consistently show that the risk of lung cancer among workers classified as having been exposed to diesel exhaust is approximatley 1.2 to 1.5 times the risk in those classified as unexposed. However, the lack of definitive exposure data for the occupationally exposed study populations precludes using the available epidemiologic data to develop quantitative estimates of cancer risk. When appropriate human information is not available, some policymakers have relied on the results of animal bioassays to estimate human risk. This document raises questions about the validity of using the rat bioassay data to characterize the potential human risk associated with ambient exposure to diesel emissions. The reason for this uncertainty is that the mechanism of lung tumor induction that appears to operate in rats continuously exposed to high concentrations of diesel exhaust and other particulate matter may not be relevant to most humans, who are exposed intermittently to levels of diesel exhaust particulate matter that are two or three orders of magnitude lower than those used in the rat bioassays. The development of unique markers of exposure to diesel emissions and a better understanding of the mechanisms of carcinogenesis would help to establish scientifically valid links between the lung cancers observed in laboratory animals and the human disease, thus improving the accuracy of cancer risk assessments.
The Health Effects Institute (HEI), established in 1980, is an independent and unbiased source of information on the health effects of motor vehicle emissions. HEI supports research on all major air pollutants. Consistent with its mission to serve as an independent source of information on the health effects of motor vehicle and other pollutants, the Institute also engages in special review and evaluation activities. Typically, HEI receives half its funds from the U.S. Environmental Protection Agency and half from 28 manufacturers and marketers of motor vehicles and engines in the United States. However, HE exercises complete autonomy in setting its research priorities, conducting evaluation and in disbursing its funds.
To learn more about HEI, please visit our home page at http://www.healtheffects.org or contact Dr. Kathleen Nauss (e-mail: knauss@healtheffects.org). Copies of the HEI Diesel Report can be obtained from The Health Effects Institute 955 Massachusetts Avenue Cambridge, MA 02139 Tel: 617-876-6700, fax 617-876-6709.